ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Status Paper on Database Issues of the Services Sector

The sustained and rapid growth of the services sector in the Indian economy has raised questions about how to accurately estimate the contribution of the sector to GDP. There are problems relating to the methodology employed on the contribution of the private sector, especially the unorganised part of the private sector. This paper outlines these problems, which were considered at the March 2006 seminar.

Status Paper on Database Issues of the Services Sector

The sustained and rapid growth of the services sector in the Indian economy has raised questions about how to accurately estimate the contribution of the sector to GDP. There are problems relating to the methodology employed on the contribution of the private sector, especially the unorganised part of the private sector. This paper outlines these problems, which were considered at the March 2006 seminar.

S L SHETTY

I The Question

A
dominant but apparently intriguing aspect of India’s growth performance has been the high and steady contribution of the services sector. According to official estimates, irrespective of the extent of the rise in GDP from agri culture and industry, GDP originating in the services sector has maintained an impressive growth rate, over 8 per cent per annum during the period 1993-94 to 2004-05 of over a decade, in contrast to an average growth of 2 per cent per annum in agriculture and a little less than 7 per cent per annum in industrial GDP. As depicted in the accompanying graphs (Graphs A, B, C and D), the annual rates of growth in the services sector have always ruled significantly above the growth rates of agriculture and industry and have also remained more stable. As a result, the share of the services sector in aggregate GDP has shown a progressive rise from 42.3 per cent in 1993-94 to around 52 per cent in 2004-05.

Two distinct questions have arisen with regard to the growth of the tertiary (i e, services) sector in the economy. The first concerns the tenability of the sector’s growth far beyond the growth of the real sectors (i e, agriculture and industry), and, that too, persistently in a secular fashion, particularly after the beginning of the 1980s. The second relates to the mis givings entertained by many regarding the quality of estimation of services sector GDP as the database depended on for the purpose is apparently weak. Implicit in the first question are both the quality of data and the economic explanation for the given relative growth of the sector. The latter question would obviously raise theoretical and historical issues and issues of the tenability of the country’s experience in this respect at this stage of its development. Before we seek to answer these substantive questions, it is felt necessary to examine the issues relating to estimation, the issues concerning the quality of the database and the nature and, if possible, the extent of data gaps embedded in the estimation of various components of the services sector. A clear insight into the working of the Indian economy and its growth scenario now hinges on the quality of the database and strength of value added estimates for different segments of the services sector. The Central Statistical Organisation (CSO) and the National Sample Survey Organisation (NSSO) have been making earnest attempts in recent years to improve the database particularly for the informal sectors. Even so, the issues raised above require renewed attention only in the spirit of offering constructive suggestions for their further improvement.

Therefore, the EPW Research Foundation (EPWRF) and the Indira Gandhi Institute of Development Research (IGIDR) organised a seminar with the essential focus on the issues concerning estimation of real GDP originating in different components of the services sector in India, the kinds of data gaps the official compilers face and what could be done to bridge the gaps. The present note was intended to serve as a theme paper for pinpointing issues for consideration and study in the research papers sought for, and deliberations at, the seminar.

II Method of Estimation and Possible Misgivings

The method of estimating GDP relating to the services sector is widely known [CSO 1999a, Kulshreshtha and Singh 1999, NSC 2002]. Briefly, of the seven major industrial categories for which the CSO regularly publishes the GDP series, four comprise the services sector, namely, (1) trade, hotels and res taurants; (2) transport, storage and communication; (3) financing, insurance, real estate and business services; and (4) community, social and personal services. Under each of these categories, there are very many sub-categories such as, “trade”, on the one hand, and “hotels and restaurants”, on the other, or “financing”, “real estate” and “business services”, or in transport, “railways”, “road transport” and “transport by other means”. Table 1 sums up their relative shares within the services sector. Amongst various sub-categories, “trade” constitutes the largest segment with above 28 per cent share in the total value added in the services sector followed by “other services” (14.5 per cent).

Again, amongst each of these categories/sub-categories, there are units operating in all the three institutional sectors: public and private corporate sectors (together constituting the organised sector) and the unorganised (household) sector. But, recognisedly, a large part of the economic activities in the services sector is carried out in the unorganised segment of the economy [Kulshreshtha et al 2002, p 13; CSO 2005, Tables 17(a) to 17(d)].

No doubt, the organised segment of the services sector is growing at a relatively faster rate but that growth is still slow, and as a result, the share of the organised segment has edged up from 47.1 per cent in 1993-94 to 53.1 per cent in 2002-03, and that of the unorganised segment has fallen only slowly in this 10-year period from 52.9 per cent to 46.9 per cent in total NDP (Table 3).

To describe the existing estimation procedure:

The estimates of GDP are prepared for the organised and the unorganised segments separately. Within the organised segment, the estimates for the public sector are based on current data and are reliable. In the case of the private corporate sector, the estimates are based on data from the Reserve Bank of India (RBI) sample

Economic and Political Weekly September 15, 2007

studies on finances of medium and large public and private limited companies. Such data become available after considerable time lag and in the absence of current data, the estimates are built up on the basis of growth in the paid-up capital. In any case, the RBI sample of companies is both small and non-random and the estimates for the universe are finally prepared by blowing up the results on the basis of paid-up capital. For the unorganised part of the activity, only the benchmark estimates are prepared using the estimates of the value added per worker (derived from the results of the follow-up surveys of the Economic Census) multiplied by the number of workers in the respective services. These benchmark estimates are thereafter carried forward with the help of suitable sector-based indices. Even in this case, the follow-up survey years are not coterminous with the benchmark year and this compounds the error in the estimates [Kulshreshtha and Singh 1999, p 137].

The quality of the database is thus admittedly weak in respect of: (1) the private corporate sector; and (2) the unorganised part of each of the services sector activities. In respect of the second category, questions have been raised on two crucial components involved in the end results: workforce estimates and estimates of value added per worker based on enterprise surveys as follow-up of the quinquennial economic censuses, which are used to work out the benchmark estimates. There is also the third component, namely, the chosen sectoral indicators used to carry the benchmark estimates forward for the recent years.

The most controversial components of the above estimation procedure are: (1) the estimates of the number of workers employed in each area of activity [CSO 1999b, Kulshreshtha et al 2000, Kumar and Sharma 2002]; and (2) results of the enterprise surveys providing estimation of average value added per worker for the corresponding areas of activity [Kulshreshtha and Singh 1999, Kar et al 2003, Subba Rao 2004, Kumar and Sharma 2004]. Various studies cited have raised serious questions on the reliability of different estimates. With a view to improving the quality of estimates, the CSO/NSSO have carried out 12 follow-up (enterprise) surveys and brought out 27 reports for different activities, mainly for covering the unorganised segment of the services sector (Appendix 1). But questions about the representative character of these sample survey results and their dependability for the purpose in view have seriously affected their reputation. For survey and estimation purposes, the unorganised segment is divided into three categories based on the size of the workforce: directory establishment (DE), employing more than five workers with at least one worker hired on a fairly regular basis; non-directory establishment (NDE), employing one to five workers with at least one hired worker; and own account enterprises (OAE), employing no hired worker. The estimates of relative shares of these categories in the workforce engaged in the economic activity as well as the respective gross value added per worker (GVAPW) estimates are used for estimating GVA in national accounting [Kulshreshtha et al 2002, p 18]. Similarly, there are alleged gaps in the estimation procedure for the private corporate sector, particularly in terms of the representative character of the sample and the variable (i e, paid-up capital) chosen for blowing up purposes.

Two sets of tabular data (Tables 2 and 3) presented here become a useful starting point to raise issues for study and debate. Table 2 depicts the sizeable increase in the share of the services sector in total NDP from 43.3 per cent in 1993-94 to 52.3 per cent in 2002-03, which as stated above

Table 2: Distribution of Net Domestic Product (NDP) by Broad Industry Groups and by Organised and Unorganised Sectors (in percentages)

Industry 1993-94 2002-03 Orgd Unorgd Total Orgd Unorgd Total

a Agriculture, forestry, fishing 3.1 50.3 32.9 2.2 40.0 23.7 b Mining, manufacturing, electricity

and construction 41.5 13.5 23.8 33.7 16.8 24.1 c Services 55.3 36.2 43.3 64.1 43.2 52.3 d Total NDP 100.0 100.0 100.0 100.0 100.0 100.0

Source: CSO (2005): National Accounts Statistics 2005, May, Government of India, p xlvi.

Table 3: Share of NDP of Each Broad Industry Group by Organised and Unorganised Sectors (in percentages)

Industry 1993-94 2002-03 Orgd Unorgd Total Orgd Unorgd Total

a Agriculture, forestry, fishing 3.5 96.5 100.0 4.1 95.9 100.0 b Mining, manufacturing, electricity

and construction 64.2 35.8 100.0 60.5 39.5 100.0 c Services 47.1 52.9 100.0 53.1 46.9 100.0 d Total NDP 36.8 63.2 100.0 43.3 56.7 100.0

Source: CSO (2005): National Accounts Statistics 2005, May, Government of India, p xlvi.

Table 1: Shares of Sub-sectors in Gross Value Added of Services Sector at 1993-94 Prices (in percentages)

Sub-sector 1950-51 1960-61 1970-71 1980-81 1990-91 2000-01 2001-02 2002-03 2003-04

1 Trade, hotels and restaurants 30.8 33.9 33.9 33.3 31.0 29.9 30.5 30.5 30.5 1a Trade 29.2 32.0 32.0 31.5 29.1 27.8 28.2 28.4 28.3 1b Hotels and restaurants 1.7 1.9 1.9 1.8 1.8 2.1 2.2 2.1 2.2

2 Transport, storage and communication 11.8 13.5 14.4 17.0 15.3 16.8 17.2 17.9 19.2 2a Railways 4.6 4.8 4.5 4.0 3.3 2.2 2.3 2.2 2.2 2b Transport by other means 5.9 7.0 7.8 10.0 9.2 8.9 8.6 8.5 8.6 2c Storage 0.3 0.2 0.2 0.3 0.2 0.1 0.1 0.1 0.1 2d Communication 1.4 1.7 2.2 2.6 2.5 5.6 6.2 7.2 8.3

3 Financing, insurance, real estate and business services 23.8 21.0 18.4 17.8 23.8 25.8 25.2 25.4 25.0 3a Banking and insurance 3.4 4.4 5.2 6.3 10.3 13.5 13.1 13.5 13.3 3b Real estate, ownership of dwellings and business services 19.7 16.1 13.1 11.5 13.6 12.3 12.2 11.9 11.7

4 Community, social and personal services 33.6 31.6 33.2 31.9 30.0 27.6 27.1 26.1 25.3 4a Public administration and defence 9.1 9.9 13.2 14.5 14.1 12.3 11.8 11.2 10.8 4b Other services 24.4 21.7 20.0 17.4 15.9 15.3 15.3 15.0 14.5

5 Total for all services 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Source: Central Statistical Organisation, National Accounts Statistics, various issues

Economic and Political Weekly September 15, 2007

Graph A: Trend Growth Rate in Real GDP Graph C: Trend Growth Rate in Real GDP Originating in Industry

(1950-51 to 2004-05) (1950-51 to 2004-05)

()

()

-6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 1950-511952-531954-551956-571958-591960-611962-631964-651966-671968-691970-711972-731974-751976-771978-791980-811982-831984-851986-871988-891990-911992-931994-951996-971998-992000-012002-032004-05 (Real Growth in Per Cent) Gross Domestic Product at Factor Cost
-4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 1950-511952-531954-551956-571958-591960-611962-631964-651966-671968-691970-711972-731974-751976-771978-791980-811982-831984-851986-871988-891990-911992-931994-951996-971998-992000-012002-032004-05 (Real Growth in Per Cent) Industry

Period Equation Growth Rate* Period Equation Growth Rate* 1950-51 to 1980-81 LnY= 11.82 + 0.0352 t 3.5821950-51 to 1980-81 LnY= 9.92 + 0.0523 t 5.377 (1261.1) (68.8) (410.6) (39.7) Adjusted R square = 0.993 Adjusted R square = 0.981 1980-81 to 2004-05 LnY= 12.82 + 0.0558 t 5.7391980-81 to 2004-05 LnY=11.41 + 0.0612 t 6.319 (1539.0) (99.5) (965.3) (77.0) Adjusted R square = 0.997 Adjusted R square = 0.995

) ()

Graph B: Trend Growth Rate in Real GDP Originating in Agriculture, Forestry and Graph D: Trend Growth Rate in Real GDP Originating in Total Service Sector Fishing (1950-51 to 2004-05) (1950-51 to 2004-05)

-15.0 -10.0 -5.0 0.0 5.0 10.0 15.0 20.0 1950-511952-531954-551956-571958-591960-611962-631964-651966-671968-691970-711972-731974-751976-771978-791980-811982-831984-851986-871988-891990-911992-931994-951996-971998-992000-012002-032004-05(Real Growth in Per Cent) Agriculture, Forestry and Fishing 0.0 2.0 4.0 6.0 8.0 10.0 12.0 1950-511952-531954-551956-571958-591960-611962-631964-651966-671968-691970-711972-731974-751976-771978-791980-811982-831984-851986-871988-891990-911992-931994-951996-971998-992000-012002-032004-05 (Real Growth in Per Cent) Total of Services
Period 1950-51 to 1980-81 Equation LnY=11.31 + 0.0215 t Growth Rate* 2.178 Period 1950-51 to 1980-81 Equation LnY= 10.50 + 0.0449 t Growth Rate* 4.601
1980-81 to 2004-05 (642.7) (22.4) Adjusted R square = 0.943 LnY=11.96 + 0.0291 t 2.953 1980-81 to 2004-05 (1630.2) (127.9) Adjusted R square = 0.998 LnY= 11.77 + 0.0708 t 7.341
(781.0) (28.2) Adjusted R square = 0.970 (991.0) (88.6) Adjusted R square = 0.996
Note: * Per cent per annum. Dividing line represents structural break

may pose questions of overestimation. On the other hand, given the absorption of a large labour force (as a residual or push factor from the real sectors), there may be misgivings that the services sector incomes are underestimated [Kansal 1992]. As shown in Table 3 and as referred to above, the share of the unorganised sector in total NDP has fallen from 52.9 per cent in 1993-94 to 46.9 per cent in 2002-03, though the perception is that the vast expansion in labour force gets absorbed in the unorganised segment of the services sector.

III Issues for Study

In view of the above, the entire methodology of estimating GDP originating in different segments of the services sector requires a close look. It is necessary to seek answers to the following issues.

First, the method of workforce estimates, applying the worker population ratio (WPR) of NSS to the population projections made by the registrar general of India (RGI), may have to be given a fresh look.

Second, are there methods of validating the results of the various enterprise surveys in regard to estimates of value added per worker? Such critical evaluation of these results, separately for trade and other segments of the services sector, will alone bring out their usefulness for national income purposes.

Third, what about the representative character of the follow-up (enterprise) surveys, the results of which are constantly found wanting insofar as the national income estimates are concerned? An examination of their inherent sampling and non-sampling errors needs to be carried out.

Fourth, there are a series of secondary indicators, which are employed for carrying the benchmark estimates forward, which may also be critically evaluated.

Fifth, a critical review is required of the method of estimation of the private corporate sector’s contribution to the GDP of the services sector. This has very often raised the question of the representative character of the sample and the variable chosen for blowing-up purposes. Has there been any progress in resolving this long-standing issue?

Sixth, financial services are a special category and the coverage of the unorganised segment in them appears weak even as this segment is said to be growing. A comprehension study of the financial services sector is called for.

Seventh, there is the most complex question of deflating the services sector output, obviously separately for each of the segments. Is there scope for revising the existing methods in this respect?

Finally, there may be other incidental issues, the examination of which may help in understanding the dynamics of the various estimation procedures.

Economic and Political Weekly September 15, 2007

Table 4: List of Enterprise Surveys as Follow-up of Economic Census, Conducted So Far by Enterprise Survey Unit of Economic Census and Surveys Division of CSO

Economic Census 1977 Report Released
Follow-up Surveys
1 Directory Manufacturing Establishment Survey (DME) 1978-79 1 DME report for ES 1978-79 – Summary results for central sample
2 Directory Establishment Survey 1979-80 Covering Sectors of Trade, Hotels, 2 DTE report for ES-1979-80 – Tables with Notes on Trade Sector
Restaurants, Transport Services and Storage and Warehousing 3 DTE repor t for ES-1979-80 – Tables with Notes on Trade Sector –
Supplement
4 Directory ES-1979-80 – Tables with Notes on Hotels and Restaurants Sector
– Summary Result – Central Sample
5 Directory ES-1979-80 – Tables with Notes on Mechanised Passenger and
Goods Transport and Services Incidental to Transport
6 Report on Directory Establishment Survey of Services, ES – 1979-80
7 Report on Storage and Warehousing, ES-1979-80
3 Enterprise Survey on Hotels, Restaurants, Transport, Storage and Ware 8 Report on Storage and Warehousing, ES-1983-84
housing and Service Sector 1983-84 9 Report on Hotels and Restaurants, ES-1983-84
4 ES-1984-85 Survey on Unorganised Manufacture: Directory Establishments 10 Report on Non-Mechanised Passenger Goods Transport, ES-1983-84
5 ES-1985-86 Survey on Directory Trade Establishments 11 Report on Mechanised Transport, ES 1983-84
6 ES-1988-89 Survey on Hotels, Restaurants and Transport 12 Report on Services Incidental to Transport, 1983-84
7 ES-1989-90 (DME) Establishment Survey 13 Report on Services Sector, ES-1983-84
8 ES-1990-91 (DTE) Establishment Survey 14 Report on DME Survey, 1984-85 – Summary and Detailed Reports
9 ES-1991-92 Service Sector Enterprise Survey 15 Report on DTE, ES – 1985-86
16 Report on Hotels and Restaurants, ES-1988-89
17 Report on Transport, ES-1988-89
18 Report on DME, ES 1989-90
19 Report on DTE, ES 1990-91
20 Report on OAE-Service Sector 1991-92
21 Report on Establishments Service Sector 1991-92
22 Report on Service Sector 1991-92
10 Enterprise Survey 1992-93 – Mining, Quarrying and Storage and Warehousing 23 Report on Storage and Warehousing, ES-1992-93
Sector. 24 Report on Mining and Quarrying, ES-1992-93
11 Enterprise Survey 1993-1994 – Hotels, Restaurants and Transport Sector 25 Report on Hotels and Restaurants, ES-1993-94
12 Directory Trade Establishment Survey 1996-97 26 Report on Transport Sector, ES-1993-94
27 Report on Trade Sector Based on Directory Trade Establishment Survey,
1996-97

Source: Ministry of Statistics and Programme Implementation (2005): Activities of the Enterprise Survey Unit of CSO, Government of India, Website:http://mospi. nic.in/t20.htm

Appendix I Activities of the Enterprise Survey Unit of CSO

  • (1) The Enterprise Survey Unit (ESU) of the Economic Census and Surveys Division of CSO is responsible for carrying out certain follow-up surveys relating to non-agricultural activities. For carrying out these surveys, economic census frame giving count of enterprises at village/block level is used for selection of sample villages/blocks to the extent feasible.
  • (2) Under the follow-up surveys, the activities of mining and quarrying, manufacturing, trade, hotels and restaurants, transport, storage and warehousing, and services have so far been covered.
  • (3) Data in these surveys are collected by the interview method. However, for enterprises maintaining books of accounts, information is collected on the basis of these records. A moving reference period is adopted for collection of most of the information to reduce the recall lapse. The survey period is generally of one year duration and this period is divided into four sub-rounds of three months each to fully capture the effect of seasonality.
  • (4) The detailed information collected in the follow-up surveys relates to employment, emoluments, fixed capital, working capital, receipts, expenses, source of finance, outstanding loans, etc.
  • (5) The ESU has so far carried out 12 follow-up surveys. In all 27 reports based on these surveys have been brought out. The list of surveys along with the reports released is shown in Table 4.
  • EPW

    Email: slshetty@vsnl.com

    References

    CSO (Central Statistical Organisation) (1999a): Working Force Estimates – 1993-94, A Methodological Note, National Accounts Division, Ministry of Planning and Programme Implementation, March.

    – (1999b): New Series on National Accounts Statistics (Base Year 1993-94), Ministry of Planning and Programme Implementation, May.

    – (2005): National Accounts Statistics, May.

    Kulshreshtha, A C and Gulab Singh (1999): ‘Services Sector in National Accounts: Methodology, Data Quality, Gaps and Possibilities of Improvement’, The Journal of Income and Wealth, IARNIW, Vol 21, No 2, July.

    Kulshreshtha, A C, Gulab Singh, Aloke Kar and R L Mishra (2000): ‘Workforce in the Indian National Accounts Statistics’, The Journal of Income and Wealth, IARNIW, Vol 22, No 2, July.

    Kulshreshtha, A C, Aloke Kar and Gulab Singh (2002): ‘Enterprise Surveys in the Improvement of Indian National Accounts Statistics’, The Journal of Income and Wealth, IARNIW, Vol 24, Nos 1 and 2, January-December.

    Kansal, S M (1992): ‘Contribution of ‘Other Services’ Sector to Gross Domestic Product in India: An Evaluation’, Economic and Political Weekly, September 19.

    Kar, Aloke, Gulab Singh and A C Kulshreshtha (2003): ‘Estimates of Value Added per Worker from Enterprise Surveys – Cross-Validating Results, Economic and Political Weekly, December 27.

    Katyal, R P and L P Rai (1999): ‘Under/Over Estimation in the New Series of National Income and Related Macro-Economic Aggregates’, The Journal of Income and Wealth, IARNIW, Vol 21, No 1, January.

    Kumar, Sanjay and N K Sharma (2002): ‘Workers in Census 2001: Some Pertinent Issues’, Economic and Political Weekly, May 4.

    – (2004): ‘Some Methodological Issues Regarding Measurement of Contribution of Unorganised Non-Agricultural Sector in the Domestic Product’, paper read at the Twenty-Fourth Conference on Income and Wealth, Organised by IARNIW jointly with Directorate of Economics and Statistics, Government of Tamil Nadu, November 5-7.

    NSC (National Statistical Commission) (2002): Report of the National Statistical Commission, Vol II, Government of India. Subba Rao, K G K (2004): ‘Estimates of Value Added Per Worker’, Economic and Political Weekly, February 14.

    Economic and Political Weekly September 15, 2007

    Dear reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here

    Comments

    (-) Hide

    EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

    Back to Top